The Official Sequential/DSI Forum

Did Polyunison Detune become more gradual with OS 2.0+?

LPF83

  • ***
  • 675
Did Polyunison Detune become more gradual with OS 2.0+?
« on: September 23, 2021, 06:22:52 PM »
Suddenly on some patches, I'm finding Polyunison2 Detune=3 to be not only usable but downright beautiful on certain patches.  Did the detuning improve in the latest OS version or am I just listening to different patches I didn't discover before?  Things are still a little too crazy too fast after Detune=4, but I never could find a good sound on anything higher than Detune=2 before. 

(disclaimer:  I have not rolled back to 1.5.2 to do a comparison and honestly would prefer not to, so hoping Pym or someone else can answer definitively)

Nothing else out there sounds like a P10 or upgraded P5 in Polyunison2 mode!
Prophet 10, Prophet 6, OB-6, Prophet 12m, Prophet Rev2-16, Toraiz AS-1, Moog SlimPhatty, Hydrasynth desktop, Korg Minilogue XDm, Roland SPD-30, Roland SPD-SX Special Edition, Roland KT-10, Maschine, Focusrite Scarlett 18i20 3rd Gen + Octopre, Strymon Pedals, Cubase Pro 11.

Re: Did Polyunison Detune become more gradual with OS 2.0+?
« Reply #1 on: September 24, 2021, 04:37:52 AM »
What you are probably hearing is the fact that poly unison now respects the P5 voice allocation mode so that each time a given note is played, it is the same pair of voices playing it. There was a bug prior to v2.0 where poly unison was always in round robin, irrespective of the voice allocation mode chosen, this meant that the detune between two voices would vary wildly, depending on which two were paired at any given time.
« Last Edit: September 24, 2021, 04:39:24 AM by Quatschmacher »

LPF83

  • ***
  • 675
Re: Did Polyunison Detune become more gradual with OS 2.0+?
« Reply #2 on: September 24, 2021, 05:20:12 AM »
What you are probably hearing is the fact that poly unison now respects the P5 voice allocation mode so that each time a given note is played, it is the same pair of voices playing it. There was a bug prior to v2.0 where poly unison was always in round robin, irrespective of the voice allocation mode chosen, this meant that the detune between two voices would vary wildly, depending on which two were paired at any given time.

Interesting.  I do tend to keep the P10 in P5 voice mode 90% of the time, and would likely not have realized it wasn't retained in PU2, since I typically only switch to PU2 for dense chords.

Thanks for this info.  Now, if we can only get a more gradual detuning amount so that the detuning amounts past 4 are more usable (without affecting whats going on now between 1-3 which I think is great).


Prophet 10, Prophet 6, OB-6, Prophet 12m, Prophet Rev2-16, Toraiz AS-1, Moog SlimPhatty, Hydrasynth desktop, Korg Minilogue XDm, Roland SPD-30, Roland SPD-SX Special Edition, Roland KT-10, Maschine, Focusrite Scarlett 18i20 3rd Gen + Octopre, Strymon Pedals, Cubase Pro 11.

Re: Did Polyunison Detune become more gradual with OS 2.0+?
« Reply #3 on: September 24, 2021, 05:33:33 AM »
What you are probably hearing is the fact that poly unison now respects the P5 voice allocation mode so that each time a given note is played, it is the same pair of voices playing it. There was a bug prior to v2.0 where poly unison was always in round robin, irrespective of the voice allocation mode chosen, this meant that the detune between two voices would vary wildly, depending on which two were paired at any given time.

Interesting.  I do tend to keep the P10 in P5 voice mode 90% of the time, and would likely not have realized it wasn't retained in PU2, since I typically only switch to PU2 for dense chords.

Thanks for this info.  Now, if we can only get a more gradual detuning amount so that the detuning amounts past 4 are more usable (without affecting whats going on now between 1-3 which I think is great).

I concur about better detuning options. I made a couple of suggestions when I first got the synth:

1. A global option which offers selectable detune amounts;

2. Detune tables (similar to tuning tables), which could be user populated.